Prescription appropriateness of anticoagulant drugs for prophylaxis of venous thromboembolism in hospitalized multimorbid older patients

The aims were to assess: the prescription prevalence of anticoagulant drugs for thromboprophylaxis (TP) in hospitalized older patients; the appropriateness of their prescription or non-prescription; the in-hospital mortality in appropriately versus non-appropriately prescribed or not prescribed patients. 4836 patients aged 65 or older, admitted to the Italian internal medicine and geriatric wards participating to the REPOSI register from 2012 to 2019 were assessed for prescription of anticoagulant drugs for TP at admission and/or during hospital stay. The Padua Prediction Score (PPS) and the IMPROVE score were used to assess the thrombotic and bleeding risk. Patients were considered to be appropriately prescribed when had PPS >= 4 and IMPROVE < 7, and appropriately not prescribed when PPS < 4. Logistic regression model was used to assess whether appropriateness was associated with in-hospital mortality. Among 4836 patients included, anticoagulants were prescribed for TP in 1233 (25.5%). In all, 4461 patients were assessable for appropriateness: 3136 (70.3%) were appropriately prescribed or non-prescribed according to their thrombotic and bleeding risk. Among 1138 patients receiving prophylaxis, only 360 (31.7%) were appropriately prescribed, while among 3323 non-prescribed patients, 2776 (83.5%) were appropriately non-prescribed. The in-hospital mortality rate was lower in patients appropriately prescribed or non-prescribed than in those inappropriately prescribed or non-prescribed (OR: 0.63; 95% CI: 0.46-0.83). In conclusion, a high prevalence of multimorbid hospitalized patients were appropriately prescribed or non-prescribed for TP with anticoagulants, appropriate non-prescription being mainly driven by a high bleeding risk. The appropriateness of prescription or non-prescription was associated with lower in-hospital mortality.